JP7802802B2 - Systems and methods for risk-based insulin delivery transformation - Google Patents

Description

CROSS-REFERENCE TO RELATED APPLICATIONS
This application claims the benefit of priority to U.S. Provisional Patent Application No. 63/145,224, entitled "SYSTEMS AND METHODS FOR RISK-BASED INSULIN DELIVERY CONVERSION," filed February 3, 2021, the contents of which are incorporated herein by reference in their entirety.

The availability and reliability of glucose time-series data has increased in recent years with the growing adoption of continuous glucose monitoring (CGM) systems and connected devices. However, despite the availability of reliable glucose data, accurate tracking of insulin and meal data, as well as optimized and effective timing of mealtime insulin boluses, continues to be a problem for many people with diabetes, resulting in poor glucose control.

Traditional diabetes management algorithms are developed iteratively over time and contain numerous modules that may have overlapping or even conflicting functionality in an attempt to provide flexibility for various user considerations and interactions.

It is with respect to these and other considerations that the various aspects and embodiments of the present disclosure are presented.

Systems and methods are provided for managing hyperglycemia and hypoglycemia by adjusting received data to provide safe and reliable control to range using automatic bolus determination, where the rate of insulin delivery is determined by the level of hyperglycemic or hypoglycemic risk. Additionally, some embodiments are directed to converting insulin delivery to a rate based on glycemic risk.

In one embodiment, the risk-based insulin delivery rate converter includes a comparator configured to receive insulin data and glucose data and to identify discrepancies between separately derived estimates of metabolic and behavioral data derived from the insulin and glucose data by quantifying the extent to which recent blood glucose measurements are inconsistent with recent insulin; and a glycemic risk assessor configured to quantify a risk of at least one of current or future hyperglycemia or hypoglycemia based on the glucose data.
and an insulin delivery monitor configured to adjust the insulin delivery rate based on data from the comparator and glycemic risk assessor.

In one embodiment, a risk-based insulin delivery rate conversion method includes receiving insulin data and glucose data at a comparison unit; identifying discrepancies between separately derived estimates of metabolic and behavioral data derived from the insulin and glucose data by quantifying the extent to which recent blood glucose measurements are inconsistent with recent insulin using a model agreement assessment unit of the comparison unit; quantifying a risk of at least one of current or future hyperglycemia or hypoglycemia based on the glucose data using a glycemic risk assessment unit; and adjusting an insulin delivery rate based on data from the comparison unit and the glycemic risk assessment unit using an insulin delivery monitoring unit.

In one embodiment, the system comprises at least one processor and a non-transitory computer-readable medium containing instructions that, when executed by the at least one processor, cause the system to receive insulin data and glucose data at a comparison unit, identify discrepancies between separately derived estimates of metabolic and behavioral data derived from the insulin and glucose data by using a model agreement assessment unit of the comparison unit to quantify the extent to which recent blood glucose measurements are inconsistent with recent insulin, quantify a current or future risk of at least one of hyperglycemia or hypoglycemia based on the glucose data using a glycemic risk assessment unit, and adjust an insulin delivery rate based on data from the comparison unit and the glycemic risk assessment unit using an insulin delivery monitor.

In one embodiment, the risk-based insulin delivery rate converter comprises a comparison unit configured to receive insulin data and glucose data and including a model agreement evaluator configured to identify discrepancies between separately derived estimates of metabolic and behavioral data derived from the insulin and glucose data by quantifying the extent to which recent blood glucose measurements are inconsistent with recent insulin; a glycemic risk assessment unit configured to quantify a risk of at least one of current or future hyperglycemia or hypoglycemia based on the glucose data; an insulin delivery monitor configured to adjust the insulin delivery rate based on data from the comparison unit and the glycemic risk assessment unit; and a reference insulin rate updater configured to determine a reference insulin rate (RIR), where the RIR is an internal reference for insulin that is considered to achieve equilibrium.

In one embodiment, a risk-based insulin delivery rate conversion method includes receiving insulin data and glucose data at a comparison unit; identifying discrepancies between separately derived estimates of metabolic and behavioral data derived from the insulin and glucose data by quantifying the extent to which recent blood glucose measurements are inconsistent with recent insulin using a model agreement assessment unit of the comparison unit; quantifying a risk of at least one of current or future hyperglycemia or hypoglycemia based on the glucose data using a glycemic risk assessment unit; adjusting the insulin delivery rate based on data from the comparison unit and the glycemic risk assessment unit using an insulin delivery monitoring unit; and determining a reference insulin rate (RIR) update unit, where the RIR is an internal reference for insulin that is believed to achieve equilibrium.

In one embodiment, the system comprises at least one processor and a non-transitory computer-readable medium containing instructions that, when executed by the at least one processor, cause the system to receive insulin data and glucose data at a comparison unit; identify discrepancies between separately derived estimates of metabolic and behavioral data derived from the insulin and glucose data by quantifying the extent to which recent blood glucose measurements are inconsistent with recent insulin using a model agreement assessment unit of the comparison unit; quantify a current or future risk of at least one of hyperglycemia or hypoglycemia based on the glucose data using a glycemic risk assessment unit; adjust an insulin delivery rate based on data from the comparison unit and the glycemic risk assessment unit using an insulin delivery monitoring unit; and determine an RIR using a reference insulin rate (RIR) update unit, the RIR being an internal reference for insulin that is believed to achieve equilibrium.

In one embodiment, the method includes receiving a plurality of inputs at a comparison component; identifying discrepancies between separately derived estimates of metabolic and behavioral data derived from the inputs; quantifying a current or future risk of hyperglycemia or hypoglycemia based on the glucose data using a glycemic risk assessment component; and adjusting an insulin delivery rate based on data from the comparison component and from the glycemic risk assessment component using an insulin delivery monitoring component.

In one embodiment, the system includes a comparison unit configured to receive a plurality of inputs and identify deviations between separately derived estimates of metabolic and behavioral data derived from the inputs; a glycemic risk assessment unit configured to quantify a current or future risk of hyperglycemia or hypoglycemia based on the glucose data; and an insulin delivery monitor configured to adjust an insulin delivery rate based on data from the comparison unit and from the glycemic risk assessment unit.

In one embodiment, the system comprises at least one processor and a non-transitory computer-readable medium containing instructions that, when executed by the at least one processor, cause the system to receive a plurality of inputs at a comparison component, identify deviations between separately derived estimates of metabolic and behavioral data derived from the inputs, quantify current or future risk of hyperglycemia or hypoglycemia based on the glucose data using a glycemic risk assessment component, and adjust an insulin delivery rate based on data from the comparison component and from the glycemic risk assessment component using an insulin delivery monitor.

This Summary introduces in a simplified form a selection of concepts that are further described below in the Detailed Description. As such, this Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.

The foregoing summary, as well as the following detailed description of exemplary embodiments, will be better understood when read in conjunction with the accompanying drawings. For the purpose of illustrating the embodiments, there are shown in the drawings exemplary configurations of the embodiments. However, the embodiments are not limited to the specific methods and instrumentalities disclosed. The drawings are as follows:

FIG. 1 is a high-level functional block diagram of one embodiment of the present invention. FIG. 1 is a block diagram of one implementation of a risk-based insulin delivery rate converter. FIG. 1 is a flow diagram of one embodiment of a method for risk-based insulin delivery rate conversion. FIG. 10 is a block diagram of one embodiment of a comparator used in risk-based insulin delivery rate conversion. FIG. 1 is a flow diagram of one embodiment of a comparison method used in risk-based insulin delivery rate conversion. FIG. 1 is a block diagram of one implementation of a glycemic risk assessor used in risk-based insulin delivery rate conversion. FIG. 1 is a flow diagram of one embodiment of a method for glycemic risk assessment used in risk-based insulin delivery rate conversion. FIG. 1 is a block diagram of one embodiment of an insulin delivery monitor for use in risk-based insulin delivery rate conversion. FIG. 1 is a flow diagram of one embodiment of a method for insulin delivery monitoring for use in risk-based insulin delivery rate conversion. 1 illustrates an exemplary computing environment in which exemplary embodiments and aspects may be implemented.

The claimed subject matter is described with reference to the drawings, wherein like reference numerals are used to refer to like elements throughout. In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the claimed subject matter. It may be evident, however, that the claimed subject matter may be practiced without these specific details. In other instances, structures and devices are shown in block diagram form to facilitate description of the claimed subject matter.

FIG. 1 is a high-level functional block diagram 100 of one embodiment of the present invention. A processor 130 communicates with an insulin device 110 and a glucose monitor 120. The insulin device 110 and the glucose monitor 120 communicate with a patient 140 to deliver insulin to the patient 140 and monitor the patient's 140 glucose level, respectively. The processor 130 is configured to perform calculations and other operations and functions described further herein. The insulin device 110 and the glucose monitor 120 may be implemented as separate devices, as a single device, within a single device, or across multiple devices. The processor 130 may be implemented locally in the insulin device 110, the glucose monitor 120, or as a standalone device (or in any combination of two or more of the insulin device 110, the glucose monitor 120, or a standalone device). The processor 130, or portions of the illustrated system, may be located remotely, such as in a server or cloud-based system.

Examples of insulin devices, such as insulin device 110, include insulin syringes, external pumps, and patch pumps that deliver insulin to a patient, typically subcutaneously. Insulin device 110 also includes devices that deliver insulin by different means, such as insulin inhalers, insulin jet injectors, intravenous infusion pumps, and implantable insulin pumps. In some embodiments, a patient uses two or more insulin delivery devices in combination, for example, injecting long-acting insulin with a syringe and using inhaled insulin before meals. In other embodiments, these devices can deliver other drugs that help control glucose levels, such as glucagon, pramlintide, or glucose-like peptide-1 (GLP-1).

Examples of glucose monitors, such as glucose monitor 120, include continuous glucose monitors that record glucose values at regular intervals, such as every 1, 5, or 10 minutes. These continuous glucose monitors can be, for example, inserted transcutaneously, fully implanted, or use electrochemical or optical sensors that measure tissue noninvasively. Examples of glucose monitors, such as glucose monitor 120, also include devices that periodically draw blood or other fluids to measure glucose, such as intravenous blood glucose monitors, microperfusion sampling, or periodic fingersticks. In some embodiments, glucose readings are provided in near real time. In other embodiments, glucose readings determined by the glucose monitor may be stored within the glucose monitor itself for subsequent retrieval.

The insulin device 110, glucose monitor 120, and processor 130 may be implemented using a variety of computing devices, such as a smartphone, a desktop computer, a laptop computer, and a tablet. Other types of computing devices may also be supported. A suitable computing device is shown in FIG. 10 as computing device 1000 and cloud-based application.

The insulin device 110, glucose monitor 120, and processor 130 may communicate over a network. The network may be of various network types, including a public switched telephone network (PSTN), a cellular network, and a packet-switched network (e.g., the Internet). While only one insulin device 110, one glucose monitor 120, and one processor 130 are shown in FIG. 1, there is no limit to the number of insulin devices, glucose monitors, and processors that may be carried. The activity monitor 150 and/or smartphone 160 may also be used to collect dietary and/or activity data from or about the patient 140 and provide the dietary and/or activity data to the processor 130.

The processor 130 may execute an operating system and one or more applications. The operating system may control which applications are executed by the insulin device 110 and/or glucose monitor 120 and how the applications interact with one or more sensors, services, or other resources of the insulin device 110 and/or glucose monitor 120.

The processor 130 may be configured and/or used to receive data from the insulin device 110 and glucose monitor 120, and in some embodiments, from the patient 140, and to perform one or more of the calculations, operations, and/or functions described further herein.

The risk-based insulin delivery transformation contemplated and described herein is applicable to any conventional diabetes management platform designed to determine and/or deliver insulin delivery rates to a patient. Applicable embodiments include, but are not limited to, conventional fully manual open-loop therapy, decision-support therapy, range-controlled automated insulin delivery (AID), control to target AID, model predictive control (MPC), linear quadratic Gaussian (LQG), proportional integral derivative (PID), and the like. In some implementations, as described further herein, an insulin delivery monitor (e.g., insulin delivery monitor 245, further described herein) adjusts the insulin delivery rate based on the deviation of the predicted metabolic state versus the actual metabolic state and the hyperglycemic risk level.

Still further, some embodiments provide an artificial pancreas (AP) algorithm that manages hyperglycemia by adjusting received data to provide safe and reliable range control using automatic bolus determination, where the rate of insulin delivery is determined by the level of hyperglycemic risk. Further embodiments may be implemented to address hypoglycemic risk. Additionally, some embodiments are directed to converting insulin delivery to a rate based on glycemic risk.

Figure 2 is a block diagram of one embodiment of a risk-based insulin delivery rate converter 230. The risk-based insulin delivery rate converter 230 includes a comparison unit 235, a glycemic risk assessment unit 240, and an insulin delivery monitoring unit 245.

Inputs to the risk-based insulin delivery rate converter 230 include continuous glucose monitoring system (CGM) data 205, other sensed input data 210, insulin data 215, user input data 220, and configuration and/or setup input data 203. External process data 225 (e.g., proposed basal rate and/or proposed bolus rate) is also input to the insulin delivery monitor portion 245 of the risk-based insulin delivery rate converter 230. Output 290 of the risk-based insulin delivery rate converter 230 includes the approved basal rate and/or the approved bolus rate.

The risk-based insulin delivery rate converter 230 operates periodically and/or on-demand to provide an approved basal rate and/or an approved bolus rate for the next time interval based on glycemic risk and model deviation. The on/off criteria for the risk-based insulin delivery rate converter 230 may be applied, for example, when the patient is initiating a bolus or based on data reliability. In some embodiments, the risk-based insulin delivery rate converter 230 operates periodically, for example, every 5 minutes, whenever a new CGM value is received, etc.

Input CGM data 205 (e.g., glucose data), other sensed input data 210, and input insulin data 215 (e.g., previously administered basal/bolus insulin from insulin on board (IOB) calculations) include respective data up to the present time (i.e., up to the present). In some embodiments, CGM data may be replaced with predicted data when CGM data is missing or unreliable for a particular time interval. User input data 220 may include data based on meals and/or exercise and/or other activities. Meals and exercise and other activities may be explicitly ignored or disallowed in some embodiments.

Additional inputs may include external process data 225, such as a suggested basal rate and/or a suggested bolus rate from an external process, which may include a pre-programmed basal profile (e.g., from an insulin pump), another AP algorithm (e.g., an AID system), patient-initiated insulin delivery (basal or bolus), etc. The systems and methods described herein convert the suggested or externally derived bolus and/or basal rate into an approved bolus and/or basal rate, as described in more detail with respect to the insulin delivery monitor 245. While arrows leading to and from certain components are shown at a high level, it should be noted that the inputs to or from any process (and resulting outputs) may be inputs to another process simultaneously, sequentially, post-processing, etc., as would be understood by one skilled in the art.

If the input includes a default basal insulin delivery profile, defined either by the patient or another system, that typically defines a minimum amount of insulin per time interval during a 24-hour continuous subcutaneous insulin infusion (CSII), it is contemplated that this profile may have a feedback loop from the risk-based insulin delivery rate converter 230 described herein. However, in some embodiments, the basal insulin delivery profile may be defined by the patient. Without wishing to be bound by theory, the patient may modify the basal rate to compensate for a missed bolus or missed meal, for example, to minimize or avoid bolus meals that may adversely affect the techniques, processes, and/or algorithms provided herein. Accordingly, the systems and methods described herein are designed to monitor (i.e., convert and approve, if necessary) proposed bolus and/or basal rates from external sources before outputting to the patient or system, or other user, entity, component, module, or device.

The comparator 235 is configured to identify discrepancies between separately derived estimates of metabolic and behavioral data (e.g., with and without CGM data) by quantifying the extent to which recent blood glucose measurements are inconsistent with recent insulin (and optionally additional data, e.g., carbohydrate records).

Figure 3 is a flow diagram of one embodiment of a method 300 for risk-based insulin delivery rate conversion. Method 300 may be performed by risk-based insulin delivery rate converter 230.

At 310, input is received, for example at the comparison unit 235. The input may include, for example, glucose data (e.g., CGM data 205), insulin data 215, other sensed input data 210, user input data 220, and/or configuration and/or setup input data 203.

At 320, a discrepancy (D) between separately derived estimates of metabolic and behavioral data derived from insulin and glucose data is identified by quantifying the extent to which recent blood glucose measurements are inconsistent with recent insulin.

At 330, the glycemic risk assessment unit 240 is used to quantify the current or future risk of hyperglycemia and/or hypoglycemia based on the glucose data.

At 340, the insulin delivery rate is modified by the insulin delivery monitor 245 based on data from the comparator 235 and the glycemic risk assessor 240.

Figure 4 is a block diagram of one embodiment of a comparison unit, such as comparison unit 235. As shown in Figure 4, comparison unit 235 includes a state estimation unit 420, a model match evaluation unit 430, and a reference insulin rate (RIR) update unit 440.

The state estimator 420 may provide an estimate of the patient's physiological and/or behavioral state based on CGM feedback, other sensed inputs, and/or user inputs.
The state estimator 420 may include a model-based state observer, such as a Kalman filter or similar, that generates estimates of the patient's physiological state (e.g., mass or concentration of glucose, insulin, or other substances in various compartments) and/or behavioral state (e.g., (current or recent) feeding or physical activity). Output from the state estimator 420 may be provided to a model match evaluator 430, which then generates one or more quantitative deviations D435. In some embodiments, the deviations may optionally be computed relative to a reference insulin rate (RIR) 425. The output may be in the form of multiple vectors/matrices containing the deviations D435 or RIRs 425, and in some embodiments, the output may optionally include a history thereof.

In some embodiments, the state estimator 420 provides one or more metabolic state estimates, which may be based on a personalized physiological model, and generates the following outputs: the adjusted estimated metabolic input, the patient's estimated metabolic state for the duration of the input data, a numerical assessment of the reliability of the estimated state, and a numerical assessment of the reliability of the adjusted estimated metabolic input. The state estimator 420 is configured to receive the (optionally filtered) extrapolation force, any extracted state, and model parameters, to the extent that a personalized physiological model is used by the estimator 420. Various estimators can be used depending on the implementation. In one implementation, the estimator 420 performs an open-loop estimation of the metabolic state from a best estimate of the metabolic state vector at the start of the time series into the future, and replays the personalized physiological model in a forward direction all the way to the end of the prediction horizon. Examples and implementations are described in detail in US Pat. No. 6,929,499, filed November 12, 2020, by inventor Stephen D. This is described in U.S. patent application Ser. No. 17/096,785, entitled "Joint State Estimation Prediction That Evaluates Differences in Predicted vs. Corresponding Received Data," by Patek.

The model match evaluator 430 may be or include a process or module that evaluates the discrepancy between two different models of metabolic and/or behavioral state for one or more state variables. In other words, the model match evaluator 430 calculates the discrepancy D435 as the difference between the state estimator variables (based on all of the available data) and what the model would have predicted for the same variables in the absence of CGM data (open-loop estimates), where the discrepancy is the difference between the two versions of the variables. In particular, if metabolism can be perfectly modeled/predicted, CGM data is not required. However, metabolism cannot be perfectly modeled/predicted, and the insulin delivery monitor 245 described herein can use the discrepancy.

Exemplary state variables include: plasma glucose concentration or mass, interstitial glucose concentration or mass, glucose in other compartments of the body, fast-acting or long-acting insulin in the subcutaneous tissue in one or more compartments, insulin in the plasma, liver, or periphery resulting from subcutaneous or intravenous infusion or from endogenous secretion, states describing the uptake, action, and clearance of insulin or glucose in various compartments of the body, pharmacokinetic and/or pharmacodynamic states associated with pharmaceuticals, states related to the absorption of carbohydrates in the diet, etc.

In one embodiment, a difference value is calculated as a measure of the extent to which recent CGM data does not match the physiological model used for state estimation. In one example, the divergence between two different open-loop predictions of metabolic and/or behavioral state is quantified as a delta, e.g., by comparison of a state observer (including a Kalman filter) to estimate other states in other compartmental models with other open-loop estimates. In one such embodiment, for each internal state x, a delta Dx is calculated (if possible) that is believed to bring the open-loop prediction into agreement with the CGM recording. Examples are described in U.S. Patent Application No. 15/580,935, filed December 8, 2017, by inventor Breton, entitled "Insulin Monitoring and Delivery System and Method for CGM-Based Fault Detection and Mitigation via Metabolic State Tracking," which is incorporated herein by reference in its entirety. For example, deltas associated with the insulin action state of the model can be calculated. However, the model match evaluator 430 may quantify these deviations D435 as variances, difference values, delta variables, etc. The continuity of the CGM signal may be considered by the model match evaluator 430, along with the values and/or trends of recent CGM data. In another embodiment, (i) all information including blood glucose (BG) and (ii) all information excluding BG are used to quantify the discrepancy between artificial intelligence (AI) and machine learning (ML) models. Other useful models include compartmental models of glucose-insulin kinetics (e.g., minimal models) that include states corresponding to insulin action, which may or may not be fine-tuned to the patient's particular physiology; Kalman filters to estimate the patient's insulin action state based on blood glucose measurements, recent insulin deliveries, and carbohydrate records; open-loop estimation of insulin action state (using only recent insulin records); etc.

The RIR updater 440 can determine an internal reference insulin rate 450. The RIR 450 may overlap with a patient-defined basal profile, but is distinct from a basal profile defined by the patient, physician, or external process, which is specifically designed for compensation for meals and other behavioral events. Rather, the RIR 450 is an internal reference for what constitutes insulin that is considered to achieve balance. The RIR 450 may be a time-averaged basal rate, adapted over time, patient-dependent, fixed, zero, programmed, learned, prescribed, etc., or a combination thereof. The RIR 450 may further be derived from total daily basal (total daily insulin or TDI), a correction factor, and/or body mass index (BMI)/weight. The RIR 450 may be updated, for example, every 5 minutes, or defined by a data acquisition rate (from a CGM). The RIR 450 can be used by the state estimator 425 to improve state estimation, BG prediction, and interpretation of deviations from the model match evaluator 430. In response, the RIR updater 440 can replace the time-varying basal rate as the basis for insulin delivery.

In some embodiments, the reference insulin rate RIR (450) is returned to the state estimation unit 420 as the insulin reference point (shown as RIR 425) in state estimation and prediction. Additionally or alternatively, in some embodiments, the deviation D (435) or the reference insulin rate RIR (450) is provided to the glycemic risk assessment unit 240 (shown in FIG. 6 as D (622 and 642) and RIR (625 and 645), respectively), in which case the RIR can serve as a reference point for interpreting past insulin delivery in quantifying risk of hypoglycemia or hyperglycemia, as an alternative to the patient's pre-programmed basal rate profile. Additionally or alternatively, in some embodiments, the deviation D (435) or reference insulin rate RIR (450) is provided to the insulin delivery monitor 245 (shown in FIG. 8 as D at 822 and RIR at 825, respectively), in which case the RIR can serve as a reference point for interpreting past and future suggested basal and/or bolus recommendations from the external process data 225, instead of the patient's pre-programmed basal rate profile. The pre-programmed basal rate profile may have time characteristics that make it inappropriate as an insulin reference, such as partial management of a normal meal with an elevated basal rate.

Figure 5 is a flow diagram of one embodiment of a comparison method 500 for use in risk-based insulin delivery rate conversion. Method 500 can be performed using the comparison unit 235.

Input is received at 510. The input may include, for example, glucose data (e.g., CGM data 205), insulin data 215, other sensed input data 210, user input data 220, and/or configuration and/or setup input data 203.

At 520, the patient's physiological and/or behavioral state is estimated based on the received input using a state estimator, such as state estimator 420. The output is provided to a model match evaluator, such as model match evaluator 430. Additionally or alternatively, the output may be provided to other components and/or modules for later use.

At 530, the discrepancy D435 between two different models of metabolic and/or behavioral state is evaluated for one or more state variables. For example, the difference between the state estimator variable and what the model would have predicted for the same variable in the absence of CGM data is calculated, where the discrepancy D is the difference between the two versions of the variable. The discrepancy D may be provided to other components and/or modules for later use.

At 540, the internal RIR is determined and provided to various components and/or modules (described further herein) for subsequent use.

Figure 6 is a block diagram of one embodiment of a glycemic risk assessor, such as glycemic risk assessor 240. The glycemic risk assessor 240 determines a hyperglycemic risk 620 and a hypoglycemic risk 640. The hyperglycemic risk 620 and the hypoglycemic risk 640 may be determined using the output of the state estimator 420. Additionally or alternatively, the hyperglycemic risk 620 and the hypoglycemic risk 640 may be determined using the model match assessor's deviations D622 (for hyperglycemic risk) and D642 (for hypoglycemic risk), respectively. Additionally or alternatively, the hyperglycemic risk 620 and the hypoglycemic risk 640 may be determined using the reference insulin rate RIR 625 and the reference insulin rate RIR 645, respectively.

The glycemic risk assessment unit 240 quantifies the current and future risk of hyperglycemia and/or hypoglycemia, respectively. The glycemic risk assessment unit 240 calculates the level of risk from inputs such as blood glucose data, insulin data, user input data, the output of the state estimator, the RIR, and the discrepancy D of the model match assessment unit, and in some embodiments may be based on predicted glucose. In some embodiments, the glycemic risk (e.g., calculated hypoglycemia and/or calculated hyperglycemia) uses prediction/state estimation. The glycemic risk assessment unit 240 may use glycemic risk spatial quantification, such as in low blood glucose index (LBGI)/high blood glucose index (HBGI), and/or the risk of hypoglycemia and/or hypoglycemia, as described in the patent application Ser. No. 09/020,493, filed on 2009/02 ... Examples and embodiments may be those described in U.S. Patek U.S. Pat. No. 10,638,981, entitled "METHOD, SYSTEM AND COMPUTER READABLE MEDIUM FOR ASSESSING ACTIONABLE GLYCEMIC RISK."

Each assessment of hyperglycemia and/or hypoglycemia may be multivariate, including a predicted BG (either for a specific range, for the entire trajectory, or for the "hurricane path"). In one embodiment, the glycemic risk assessment component 240 may include a delta as described by Breton, such as the example described in U.S. Patent Application No. 15/580,935, filed December 8, 2017, to inventor Breton, entitled "Insulin Monitoring and Delivery System and Method for CGM-Based Fault Detection and Mitigation via Metabolic State Tracking," which is incorporated herein by reference in its entirety. The glycemic risk assessor 240 may be configured to assess hyperglycemic risk alone or in combination with hypoglycemic risk, such as that described in U.S. Patent Application No. 14/659,500, filed March 16, 2015, by inventor Rack-Gomer, entitled "GLYCEMIC URGENCY ASSESSMENT AND ALERT SINTERFACE," which is incorporated herein by reference in its entirety. The fit of the risk function may be parameterized. The glycemic risk assessor 240 may be configured to assess hyperglycemic risk alone or in combination with hypoglycemic risk, such as that described in U.S. Patent Application No. 14/659,500, filed March 16, 2015, by inventor Rack-Gomer, entitled "GLYCEMIC URGENCY ASSESSMENT AND ALERT SINTERFACE," which is incorporated herein by reference in its entirety. The fit of the risk function may be parameterized. Normalized risk, as described in U.S. Patek, U.S. Pat. No. 10,638,981, entitled "METHOD, SYSTEM AND COMPUTER READABLE MEDIUM FOR ASSESSING ACTIONABLE GLYCEMIC RISK," allows for parameterization of the shape of the risk function in a more natural way. Exemplary risk-based windows could be, for example, over 5 minutes, over 30 minutes, a basal/bolus combination, a function of time, etc.

Figure 7 is a flow diagram of one embodiment of a method 700 for glycemic risk assessment for use in risk-based insulin delivery rate conversion. Method 700 can be performed using the glycemic risk assessment unit 240.

At 710, inputs such as blood glucose data, insulin data, user input data, state estimator output, RIR, and/or model match evaluator deviation D are received.

At 720, the current and/or future risk of hyperglycemia is determined (e.g., quantified).

At 730, the current and/or future risk of hypoglycemia is determined (e.g., quantified).

At 740, the risk is output to an insulin delivery monitor (e.g., insulin delivery monitor 245), the patient, a physician, or other medical professional or administrator.

Figure 8 is a block diagram of an embodiment of an insulin delivery monitor, such as insulin delivery monitor 245. Insulin delivery monitor 245 includes a nominal insulin planning component 820 and a monitor 840.

The insulin delivery monitor 245 adjusts the insulin delivery rate based on data from the comparator 235 and the glycemic risk assessor 240. The insulin delivery monitor 245 also considers suggested bolus and/or basal rates from external processes, if available. Often, suggested (basal or bolus) insulin rates are available, for example, from traditional fully manual open-loop therapy (CSII basal insulin profile), decision-support therapy (recommended algorithm), range-controlled automatic insulin delivery (AID), target-controlled AID, MPC, LQG, PID, etc. However, the systems and methods described herein can function within fully standalone algorithms and in some implementations.

Depending on the implementation, the insulin delivery monitor 245 may include insulin intensification (increasing the rate) based on hyperglycemia risk, insulin attenuation (decreasing the rate) based on hypoglycemia risk, or both. In some embodiments, the insulin delivery monitor 245 calculates the insulin rate for a time window during which the required amount of insulin will be delivered, the time window being determined from the level of glycemic risk, e.g., hyperglycemia risk or hypoglycemia risk. In some embodiments, the insulin rate is calculated based on the comparator 235, e.g., the model match evaluator 430.

In one embodiment, the insulin delivery monitor 245 adapts the proposed basal rate to the approved basal rate by adapting the proposed value according to the risk of hyperglycemia to ensure that BG remains below an upper envelope of acceptable values, where the upper envelope is a function of time (e.g., may be the time of day or may vary with respect to other parameters), as further described herein.

The nominal insulin planning component 820 takes into account the risk of hyperglycemia and/or hypoglycemia (and optionally uses the estimated impairment state and RIR 825) to determine a future insulin target trajectory, which is converted into suggested basal and/or bolus rates. The nominal insulin planning component 820 can function as a supplemental layer for an existing algorithm or within a stand-alone algorithm.

The nominal insulin planning unit 820 determines the amount of insulin required to minimize the deviation determined by the model match evaluation unit 430 of the comparison unit 235. This amount is determined based on the model fit calculations described in U.S. Patent Application Publication No. 2019/0254595 A1, filed December 8, 2017, and published by inventor Marc D. This may be a standard amount of insulin needed, such as the insulin that should be on board (ISOB) (residual insulin requirement) as described in U.S. Patent Application No. 15/580,935 to Breton, entitled "Insulin Monitoring and Delivery System and Method for CGM-Based Fault Detection and Mitigation via Metabolic State Tracking," or may be provided in terms of future plasma insulin or other physiological considerations.

In one exemplary embodiment (e.g., as described in U.S. Patent Application No. 15/580,935, incorporated herein by reference in its entirety), the condition being evaluated is IOB vs. ISOB, based on how much insulin is expected to be required to return the patient to an upper BG envelope curve, where the upper BG envelope is a time-of-day curve, e.g., curve values are higher during the day (e.g., 160 mg/dl) and drop (e.g., 120 mg/dl) at night. In some embodiments, the upper BG envelope is calculated based on a current estimate of BG, designed to allow BG to drop to a final value during a time window. In some embodiments, the systems and methods described herein ensure a meaningful response to hyperglycemic risk by imposing a maximum curve value to address the issue of sometimes inadequate response to hyperglycemia. Notably, while the upper envelope is used as a target by ISOB, this is not the same as the target of the control algorithm for fine-tuning insulin delivery. In some embodiments, the nominal insulin planning component 820 calculates ISOB based on a target defined by an upper BG envelope curve that is generated on-demand based on estimated BG. In some embodiments, the envelope is determined from a sleep profile, although this is not a requirement and may not actually be done in a particular implementation. In some embodiments, ISOB may be calculated as a function of both the upper and lower BG envelopes; for example, ISOB may be calculated to achieve a BG somewhere between a hyperglycemic upper envelope curve value and a low BG envelope according to an insulin cutoff threshold or logic.

In some embodiments, insulin may be expressed directly in terms of subcutaneous insulin delivery (see, e.g., ISOB described in U.S. Patent Application No. 15/580,935, incorporated herein by reference in its entirety) relative to a user-provided basal rate profile or relative to a reference insulin rate (RIR). In some embodiments, the output of the nominal insulin planning component may be based on the BG upper envelope curve or other mechanisms for optimizing the patient's theoretical insulin trajectory.

The time window (sometimes called the "rate window") used to determine the insulin delivery rate (which will indicate the level of concordance or discordance) may be a function of the risk of glycemia (e.g., hyperglycemia or hypoglycemia) and is therefore variable. For example, if the risk of hyperglycemia is high, the entire amount of insulin needed may be delivered as fast as possible, i.e., as a bolus.

The monitoring unit 840 may be combined with or separate from the nominal insulin planning unit 820. The monitoring unit 840 reconciles suggested basal rates (and optionally suggested bolus rates) from an external source (e.g., external process data 225) with the insulin needs identified by the nominal insulin planning unit 820 to determine the approved basal rate (and/or bolus) for the next periodic update. U.S. Patent Application No. 15/580,935, incorporated herein by reference in its entirety, describes an embodiment in which insulin needs are determined by calculating ISOB and comparing it to IOB. However, other methods of determining insulin needs may be used.

In addition to inputs from the original inputs, the monitor 840 may process the output of the state estimator 420, the output of the model match evaluator 430 (i.e., the deviation D), the output of the RIR updater 440 (i.e., the RIR 450), and may further include inputs from an externally derived process describing basal insulin and, optionally, bolus insulin (i.e., external process data 225). Thus, the monitor 840 may be useful for coordinating external processes with the systems and methods described herein for insulin planning.

In some embodiments, the insulin delivery monitor 245 may convert the bolus recommendation into a combination of bolus and basal, e.g., an amount delivered at a maximum rate and an amount delivered as an elevated basal rate over a period of time. The conversion can be based on the state of the system and glycemic risk and can be fed back to previous steps and/or modules of the methods and/or systems described herein. In some embodiments, the conversion can be informed by when the next decision can be made. In some embodiments, the risk-based insulin delivery rate converter 230 takes the output of any open-loop or closed-loop artificial pancreas algorithm designed to generate a rate of insulin delivery and converts that rate into a blend of a basal rate and an individual bolus. In some embodiments, the individual (correction) bolus is coordinated with the basal rate based on hyperglycemic risk, i.e., the insulin delivery monitor 245 converts the recommended correction bolus to a rate where the rate window is calculated as a function of hyperglycemic risk (rather than a fixed rate window such as 30 minutes).

In one example, predicted blood glucose is used to calculate hyperglycemic risk, which is used by the model match evaluator 430 to quantify the variance between glycemic and/or insulin states, where the greater the risk of hyperglycemia, the shorter the rate window. In other words, at the highest level of hyperglycemia, the required amount of insulin is delivered as a discrete bolus. Thus, the rate window is variable, such that the higher the calculated risk of hyperglycemia, the closer the rate window is to 5 minutes (or any periodic rate of data acquisition refresh and/or controller update). As an example, if the risk-based insulin delivery rate converter 230 calculates that the difference between ISOB and IOB is 3 units, this may be delivered in 5 minutes at a high level of hyperglycemic risk, but over 30 minutes at a low level of hyperglycemic risk. Specifically, the pre-intervention hyperglycemic risk is used to convert ISOB to a rate of insulin delivery that will be applied until the next controller update, resulting in ISOB being delivered as a discrete bolus at a high hyperglycemic risk. However, in some embodiments, the conversion of ISOB to speed may be based on knowledge of predicted BG both pre- and post-intervention. In contrast to standard model predictive control (MPC), the speed value is a modified ISOB value, which is not the result of optimization.

Although the above example describes the use of hyperglycemic risk, the conversion of ISOB to rate may be based on both hyperglycemic and hypoglycemic risk information.

The rate window can be the denominator for converting the individual bolus to a rate based on hyperglycemic risk, as shown in the equation below.
Insulin delivery rate = (amount of insulin required based on level of concordance)/(risk-based time window during which the amount of insulin required will be delivered).

In this case, the resulting recommendation from the monitoring unit 840 can be large enough to achieve the effect of individual corrections and/or meal boluses, or small enough to include a low basal delivery rate.

In some embodiments, the aggressiveness of the insulin delivery monitor 245 may be constrained based on an assessment of the patient's total daily insulin requirement (TDI). For example, parameters needed to calculate an appropriate response to the difference between IOB and ISOB may be constrained as a function of TDI. Ongoing modifications of the TDI adjust how aggressive the nominal insulin planning component is allowed to be. A saturation value for the correction factor may be used as a separate check on how aggressive the control algorithm is allowed to be. Here, limits on the correction factor may be implemented.

Output 290 from insulin delivery monitor 245 includes the approved basal rate and, optionally, the approved bolus rate. Output 290 may also include a message sent to the patient, a physician or other medical professional or administrator, a display, a computing device, etc. For example, a predicted BG trajectory may be displayed along with an explanation of the uncertainty. Recommended values or amounts of insulin delivery may be provided or described for specific time intervals and/or with respect to various conditions (e.g., "if," "when," "based on," "time in range without meal notification," etc.).

Figure 9 is a flow diagram of one embodiment of a method 900 for insulin delivery monitoring for use in risk-based insulin delivery rate conversion. Method 900 can be performed using the insulin delivery monitor unit 245.

At 910, inputs such as current and/or future risk of hyperglycemia and/or hypoglycemia, the output of the state estimator 420, the output of the model match evaluator 430 (i.e., the deviation D), the output of the RIR updater 440 (i.e., the RIR 450), and inputs from an externally derived process describing basal insulin and, optionally, bolus insulin are received.

At 920, a future insulin target trajectory is determined.

At 930, the nominal insulin planning component is used to determine the amount of insulin required to minimize the deviation D from the model fit assessment component 430 and/or minimize the risk of hyperglycemia.

At 940, the proposed basal rate and/or proposed bolus rate are adjusted with the insulin needs identified by the nominal insulin planning component to determine the approved basal rate and/or approved bolus rate.

At 950, the approved basal rate and/or approved bolus rate are output, for example, to the delivery device, the patient, a physician or other medical professional or administrator, a display device, a computing device, etc.

Example 1 - Monitoring the implementation of conventional insulin pump therapy

In one embodiment, the systems and methods described herein are operatively used in conjunction with an insulin pump therapy system (external process) having a user-programmed basal rate profile and a functional pre-meal insulin bolus calculated using carbohydrate estimates, carbohydrate ratios, correction factors, and IOBs. In this example, the system/method operates as follows:

The comparator 235 quantitatively adjusts the open-loop and CGM-based estimates of the current metabolic state vector in various ways, including one or more of the following: by attributing the level of agreement (or discrepancy) to a failure of insulin delivery (e.g., pump blockage) → setting a pump failure state estimate, by attributing agreement/disagreement to an unexpectedly low/high "insulin action" → recognizing the fact that the insulin sensitivity parameter is too low/too high, and/or by attributing the quantified agreement/disagreement to an unannounced meal (or a meal with a higher carbohydrate content than the patient approved) → incrementally adapting the model agreement evaluator 430's deviation (D) for insulin action.

The glycemic risk assessment unit 240 estimates the applicable quantitative values of hyperglycemic risk and/or hypoglycemic risk over the specified planning period. Note that this does not assume any further intervention from the user.

The insulin delivery monitor 245 (knowing the user-programmed basal rate profile from the operably connected insulin delivery device) may optionally look at the current bolus request from the patient, estimate the effect of the basal rate profile over the specified planning period without assuming future intervention from the patient, and decide to modify the current bolus request (if any) or issue an unrequested insulin bolus, modify the basal rate profile for the duration of the planning period, and/or specify that a bolus be delivered at some future time during the planning period.

In one exemplary situation (set of conditions) according to this embodiment, when the preprogrammed basal profile is elevated relative to the patient's fasting basal profile (or RIR), the elevated basal rate may represent the user's attempt to partially cover an unannounced meal with basal insulin delivery. In this case, the monitor serves to accelerate the effect of the elevated basal rate by converting a portion of it into a separate bolus.

In another exemplary situation (set of conditions) according to this embodiment, if the comparison unit 235 can determine that either (i) unannounced/underestimated carbohydrates are present or (ii) reduced insulin sensitivity is the most likely explanation for the model mismatch, the glycemic risk assessment unit 240 can estimate a clinically significant increased risk of hyperglycemia R, and if the user has recently specified an individual bolus B, the insulin delivery monitor 245 can either: now deliver a bolus equal to B plus a percentage F of the total amount of insulin associated with the pre-programmed basal rate profile for the specified planning period T (the percentage is calculated as a function of the estimated risk of hyperglycemia R (e.g., F=k*R/(1+k*R), where k is a parameter)), and decide to deliver the remaining percentage of insulin (1-F) as a new, reduced, temporary basal rate for the specified planning period T; or, at the user's discretion, deliver only B and wait for a future opportunity for the monitor to preemptively convert basal insulin to a bolus (this can depend on various factors such as data reliability).

In yet another exemplary situation (set of conditions) according to this embodiment, if the comparison unit 235 estimates that unexpectedly high insulin action (suggesting momentarily high insulin sensitivity) is the most likely explanation for the model mismatch, the glycemic risk assessment unit 240 estimates a clinically significant elevated risk of hypoglycemia R, and if the user has recently entered an individual bolus B≧0, the insulin delivery monitor unit 245 may set a temporary basal rate for the specified planning period based on the risk. For example, based on a desire to achieve a specific IOB, knowing the user's bolus B, the basal rate can be set to achieve that IOB within a specified time frame, where both the target IOB and the time frame are calculated as a function of the estimated risk of hypoglycemia. Additionally or alternatively, the insulin delivery monitor unit 245 may warn the user about the bolus, suggesting that the bolus may worsen the risk of hypoglycemia in the absence of additional carbohydrates.

Example 2 - Monitoring an Automated Insulin Delivery (AID) Algorithm

In this second exemplary embodiment, the systems and methods described herein are operatively used in conjunction with an automatic insulin delivery therapy system (external process) that includes automatic adjustment of basal rates and/or automatic insulin boluses, regardless of whether the patient has an opportunity to request a bolus. The comparator 235 quantitatively adjusts the open-loop and CGM-based estimates of the current metabolic state vector as described in Example 1 and further based on the estimated RIR.

The glycemic risk assessment unit 240 estimates the applicable quantitative value of hypoglycemic risk over the specified planning period (note: assumes no further intervention from the user).

Based on the patient's RIR, and depending on the basal rate recommendation and/or insulin bolus request/recommendation higher than the RIR, the insulin delivery monitor 245 may decide to modify the current bolus request/recommendation (if any) or introduce a new bolus, modify the AID basal rate recommendation, and/or specify that a bolus be delivered at some future time within a specified planning period, assuming no future intervention from the patient.

In one exemplary situation (set of conditions) according to this second embodiment, if the comparator 235 determines that either (i) unannounced/underestimated carbohydrates are present or (ii) reduced insulin sensitivity is the most likely explanation for the model mismatch, the glycemic risk profiler estimates a clinically significant elevated risk of hyperglycemia R, and in the absence of a bolus recommendation or request, the insulin delivery monitor 245 may now deliver a bolus equal to a percentage F of the total amount of insulin associated with the AID-recommended basal rate profile for the specified planning period T (the percentage calculated as a function of the estimated risk of hyperglycemia R (e.g., F = k * R / (1 + k * R), where k is a parameter)) and deliver the remaining percentage of insulin associated with the AID-recommended basal rate recommendation (1 - F) as a new reduced basal rate. Additionally or alternatively, the bolus introduced above may be calculated as a function of the difference between the AID-recommended basal rate and the patient's RIR.

In another exemplary situation (set of conditions) according to this second embodiment, if the comparison unit 235 recognizes that either (i) unannounced/underestimated carbohydrates are present or (ii) decreased insulin sensitivity is the most likely explanation for the model mismatch, the glycemic risk assessment unit 240 estimates a clinically significant increased risk of hyperglycemia R, and if the user specifies only an individual bolus B, the insulin delivery monitoring unit 245 determines to now deliver a bolus equal to a certain percentage F of the total amount of insulin associated with the AID-recommended basal rate profile for the specified planning period T plus B (the certain percentage is calculated as a function of the estimated risk of hyperglycemia R, for example, F = k * R / (1 + k * R), where k is a parameter), and to deliver the remaining percentage of insulin associated with the AID-recommended basal rate value (1 - F) as a new reduced basal rate. Alternatively, the insulin delivery monitor 245 may deliver only B at the user's discretion and wait for a future opportunity for the monitor to preemptively convert basal insulin to a bolus, which may be based on data reliability or fail-safe features, for example.

In yet another exemplary situation (set of conditions) according to this second embodiment, if the comparison unit 235 recognizes that unexpectedly high insulin action (suggesting momentarily high insulin sensitivity) is the most likely explanation for the model mismatch, the glycemic risk assessment unit 240 estimates a clinically significant elevated risk of hypoglycemia R, and if the user specifies only an individual bolus B ≥ 0, the insulin delivery monitor unit 245 determines to set a temporary basal rate for the specified planning period. For example, based on a desire to achieve a specific IOB, if the user's bolus B is known, the basal rate can be set to achieve that IOB within a specified time frame, where both the target IOB and the time frame are calculated as a function of the estimated risk of hypoglycemia. If it turns out that a reduced basal rate was necessary (e.g., to compensate for unannounced carbohydrates), the difference can be introduced later as a compensated elevated basal rate or an individual bolus. Additionally or alternatively, the monitor may decide to leave bolus B unchanged, but may display/warn the user about the bolus unless B=0, suggesting that the bolus may exacerbate the risk of hypoglycemia in the absence of additional carbohydrates.

Figure 10 illustrates an exemplary computing environment in which example embodiments and aspects may be implemented. The computing device environment is only one example of a suitable computing environment and is not intended to suggest any limitation as to the scope of use or functionality.

Numerous other general-purpose or special-purpose computing device environments or configurations may also be used. Examples of well-known computing devices, environments, and/or configurations that may be suitable for use include, but are not limited to, personal computers, server computers, handheld or laptop devices, multiprocessor systems, microprocessor-based systems, networked personal computers (PCs), minicomputers, mainframe computers, embedded systems, distributed computing environments that include any of the above systems or devices, and the like.

Computer-executable instructions such as program modules executed by a computer may be used. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. Distributed computing environments may be used in which tasks are performed by remote processing devices that are linked through a communications network or other data transmission medium. In a distributed computing environment, program modules and other data may be located in both local and remote computer storage media, including memory storage devices.

With reference to FIG. 10, an exemplary system for implementing aspects described herein includes a computing device, such as computing device 1000. In its most basic configuration, computing device 1000 typically includes at least one processing unit 1002 and memory 1004. Depending on the exact configuration and type of computing device, memory 1004 may be volatile (such as random access memory (RAM)), non-volatile (such as read-only memory (ROM) or flash memory), or some combination of the two. This most basic configuration is indicated in FIG. 10 by dashed line 1006.

Computing device 1000 may have additional features/functionality. For example, computing device 1000 may include additional storage (removable and/or non-removable) including, but not limited to, magnetic or optical disks or tape. Such additional storage is illustrated in FIG. 10 by removable storage 1008 and non-removable storage 1010.

Computing device 1000 typically includes a variety of computer-readable media. Computer-readable media can be any available media that can be accessed by device 1000, including both volatile and nonvolatile media, removable and non-removable media.

Computer storage media include volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer-readable instructions, data structures, program modules, or other data. Memory 1004, removable storage 1008, and non-removable storage 1010 are all examples of computer storage media. Computer storage media include, but are not limited to, RAM, ROM, electrically erasable program read-only memory (EEPROM), flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and that can be accessed by computing device 1000. Any such computer storage media may be part of computing device 1000.

Computing device 1000 may contain communications connections 1012 that allow the device to communicate with other devices. Computing device 1000 may also have input devices 1014, such as a keyboard, mouse, pen, voice input device, touch input device, etc. Output devices 1016, such as a display, speakers, printer, etc., may also be included. All of these devices are well known in the art and need not be discussed at length here.

In one embodiment, the risk-based insulin delivery rate converter comprises a comparison unit configured to receive insulin data and glucose data and including a model agreement evaluator configured to identify discrepancies between separately derived estimates of metabolic and behavioral data derived from the insulin and glucose data by quantifying the extent to which recent blood glucose measurements are inconsistent with recent insulin; a glycemic risk assessment unit configured to quantify a risk of at least one of current or future hyperglycemia or hypoglycemia based on the glucose data; and an insulin delivery monitor configured to adjust the insulin delivery rate based on data from the comparison unit and the glycemic risk assessment unit.

In one embodiment, a risk-based insulin delivery rate conversion method includes receiving insulin data and glucose data at a comparison unit; identifying discrepancies between separately derived estimates of metabolic and behavioral data derived from the insulin and glucose data by quantifying the extent to which recent blood glucose measurements are inconsistent with recent insulin using a model agreement assessment unit of the comparison unit; quantifying a risk of at least one of current or future hyperglycemia or hypoglycemia based on the glucose data using a glycemic risk assessment unit; and adjusting an insulin delivery rate based on data from the comparison unit and the glycemic risk assessment unit using an insulin delivery monitoring unit.

In one embodiment, the system comprises at least one processor and a non-transitory computer-readable medium containing instructions that, when executed by the at least one processor, cause the system to receive insulin data and glucose data at a comparison unit, identify discrepancies between separately derived estimates of metabolic and behavioral data derived from the insulin and glucose data by using a model agreement assessment unit of the comparison unit to quantify the extent to which recent blood glucose measurements are inconsistent with recent insulin, quantify a current or future risk of at least one of hyperglycemia or hypoglycemia based on the glucose data using a glycemic risk assessment unit, and adjust an insulin delivery rate based on data from the comparison unit and the glycemic risk assessment unit using an insulin delivery monitor.

Implementations may include some or all of the following features. The computer-readable medium further includes instructions that, when executed by at least one processor, cause the system to use the model match evaluator to evaluate the discrepancy between two different models of metabolic or behavioral state and provide the discrepancy as an output for later use. The computer-readable medium further includes instructions that, when executed by the at least one processor, cause the system to use the model match evaluator to quantify the discrepancy between two different open-loop predictions of the metabolic or behavioral state as a variance. The computer-readable medium further includes instructions that, when executed by the at least one processor, cause the system to use the state estimator of the comparator to estimate at least one of the patient's physiological or behavioral state based on at least one of continuous glucose monitoring system (CGM) feedback, other sensed input, or user input and provide an output to the model match evaluator. The state estimate is used by the model match evaluator. The computer-readable medium further includes instructions, when executed by the at least one processor, that cause the system to assess a hyperglycemic risk by the glycemic risk assessment unit, and the hyperglycemic risk is used to adjust a time window used to calculate an insulin rate by the insulin delivery monitoring unit. The insulin delivery monitoring unit considers at least one of a suggested bolus rate or a basal rate from an external process. The insulin delivery monitoring unit calculates an insulin rate for a time window during which the required amount of insulin will be delivered. The time window is determined from the level of glycemic risk quantified by the glycemic risk assessment unit. The computer-readable medium further includes instructions, when executed by the at least one processor, that cause the system to determine, by an insulin planning unit of the insulin delivery monitoring unit, an amount of insulin needed to minimize the deviation determined by the comparison unit. The computer-readable medium further includes instructions, when executed by the at least one processor, that cause the system to adjust, by a monitoring unit of the insulin delivery monitoring unit, a suggested basal rate from an external source with the insulin need identified by the insulin planning unit to determine an approved basal rate for the next periodic update. The computer-readable medium further includes instructions that, when executed by the at least one processor, cause the system to convert the approved basal rate into a blend of basal rate and individual boluses using the insulin delivery monitor.

In one embodiment, the risk-based insulin delivery rate converter comprises a comparison unit configured to receive insulin data and glucose data and including a model agreement evaluator configured to identify discrepancies between separately derived estimates of metabolic and behavioral data derived from the insulin and glucose data by quantifying the extent to which recent blood glucose measurements are inconsistent with recent insulin; a glycemic risk assessment unit configured to quantify a risk of at least one of current or future hyperglycemia or hypoglycemia based on the glucose data; an insulin delivery monitor configured to adjust the insulin delivery rate based on data from the comparison unit and the glycemic risk assessment unit; and a reference insulin rate updater configured to determine a reference insulin rate (RIR), where the RIR is an internal reference for insulin that is considered to achieve equilibrium.

In one embodiment, a risk-based insulin delivery rate conversion method includes receiving insulin data and glucose data at a comparison unit; identifying discrepancies between separately derived estimates of metabolic and behavioral data derived from the insulin and glucose data by quantifying the extent to which recent blood glucose measurements are inconsistent with recent insulin using a model agreement assessment unit of the comparison unit; quantifying a risk of at least one of current or future hyperglycemia or hypoglycemia based on the glucose data using a glycemic risk assessment unit; adjusting the insulin delivery rate based on data from the comparison unit and the glycemic risk assessment unit using an insulin delivery monitoring unit; and determining a reference insulin rate (RIR) update unit, where the RIR is an internal reference for insulin that is believed to achieve equilibrium.

In one embodiment, the system comprises at least one processor and a non-transitory computer-readable medium containing instructions that, when executed by the at least one processor, cause the system to receive insulin data and glucose data at a comparison unit; identify discrepancies between separately derived estimates of metabolic and behavioral data derived from the insulin and glucose data by quantifying the extent to which recent blood glucose measurements are inconsistent with recent insulin using a model agreement assessment unit of the comparison unit; quantify a current or future risk of at least one of hyperglycemia or hypoglycemia based on the glucose data using a glycemic risk assessment unit; adjust an insulin delivery rate based on data from the comparison unit and the glycemic risk assessment unit using an insulin delivery monitoring unit; and determine an RIR using a reference insulin rate (RIR) update unit, the RIR being an internal reference for insulin that is believed to achieve equilibrium.

Implementations may include some or all of the following features: The RIR update unit is included within the comparison unit. The RIR is used by the comparison unit. The glycemic risk assessment unit is configured to receive the RIR and use the RIR to quantify a risk of at least one of current or future hyperglycemia or future hypoglycemia. The insulin delivery monitoring unit is configured to receive the RIR and use the RIR to determine a future insulin target trajectory and an amount of insulin needed to minimize the deviation. The insulin delivery monitoring unit is further configured to receive deviation data and use the deviation data to determine a future insulin target trajectory and an amount of insulin needed to minimize the deviation. The computer-readable medium further includes instructions that, when executed by the at least one processor, cause the system to use the model match assessment unit to assess a deviation between two different models of a metabolic or behavioral state and provide the deviation as an output for later use. The computer-readable medium further includes instructions, when executed by at least one processor, that cause the system to quantify, as variance, the deviation between two different open-loop predictions of a metabolic or behavioral state using a model match evaluator. The computer-readable medium further includes instructions, when executed by at least one processor, that cause the system to use a state estimator in the comparator to estimate at least one of a patient's physiological or behavioral state based on at least one of continuous glucose monitoring (CGM) feedback, other sensed input, or user input, and provide an output to the model match evaluator. The state estimate is used by the model match evaluator. The computer-readable medium further includes instructions, when executed by the at least one processor, that cause the system to assess hyperglycemic risk using a glycemic risk assessor, which assesses hyperglycemic risk, and the hyperglycemic risk is used to adjust a time window used to calculate an insulin rate by the insulin delivery monitor. The insulin delivery monitor considers at least one of a suggested bolus rate or a basal rate from an external process. The insulin delivery monitor calculates an insulin rate for a time window during which the required amount of insulin will be delivered. The time window is determined from the level of glycemic risk quantified by the glycemic risk assessment unit. The computer-readable medium further includes instructions that, when executed by the at least one processor, cause the system to determine, by an insulin planning unit of the insulin delivery monitoring unit, the amount of insulin needed to minimize the deviation determined by the comparison unit. The computer-readable medium further includes instructions that, when executed by the at least one processor, cause the system to cause, by a monitoring unit of the insulin delivery monitoring unit, to adjust a suggested basal rate from an external source with the insulin need identified by the insulin planning unit to determine an approved basal rate for the next periodic update. The computer-readable medium further includes instructions that, when executed by the at least one processor, cause the system to convert the approved basal rate into a blend of basal rate and individual boluses using the insulin delivery monitoring unit.

In one embodiment, the method includes receiving a plurality of inputs at a comparison component; identifying discrepancies between separately derived estimates of metabolic and behavioral data derived from the inputs; quantifying a current or future risk of hyperglycemia or hypoglycemia based on the glucose data using a glycemic risk assessment component; and adjusting an insulin delivery rate based on data from the comparison component and data from the glycemic risk assessment component using an insulin delivery monitoring component.

In one embodiment, the system includes a comparison unit configured to receive a plurality of inputs and identify deviations between separately derived estimates of metabolic and behavioral data derived from the inputs; a glycemic risk assessment unit configured to quantify a current or future risk of hyperglycemia or hypoglycemia based on the glucose data; and an insulin delivery monitor configured to adjust an insulin delivery rate based on data from the comparison unit and from the glycemic risk assessment unit.

In one embodiment, the system comprises at least one processor and a non-transitory computer-readable medium containing instructions that, when executed by the at least one processor, cause the system to receive a plurality of inputs at a comparison component, identify deviations between separately derived estimates of metabolic and behavioral data derived from the inputs, quantify current or future risk of hyperglycemia or hypoglycemia based on the glucose data using a glycemic risk assessment component, and adjust an insulin delivery rate based on data from the comparison component and data from the glycemic risk assessment component using an insulin delivery monitor.

Implementations may include some or all of the following features: The input includes at least one of glucose data, insulin data, sensed input data, or user-input data. Identifying the discrepancy includes quantifying the extent to which recent blood glucose measurements are inconsistent with recent insulin. The computer-readable medium, when executed by at least one processor, further includes instructions that cause the system to, in the comparison unit, estimate a physiological or behavioral state of the patient based on the received input using a state estimator, cause a model agreement assessor to provide an output, evaluate a discrepancy between two different models of metabolic or behavioral state for one or more state variables, and output the discrepancy. Evaluating the discrepancy includes calculating a difference between a state estimator variable and what the model would have predicted for the same variable in the absence of continuous glucose monitoring (CGM) data, the discrepancy being the difference between the two versions of the variable. The computer-readable medium, when executed by at least one processor, further includes instructions that cause the system to determine an internal reference insulin rate (RIR) and output the RIR. The computer-readable medium further includes instructions that, when executed by the at least one processor, cause the system to determine, in the glycemic risk assessment unit, at least one of a current or future risk of hyperglycemia, determine a current or future risk of hypoglycemia, and output the current or future risk of hyperglycemia and the current or future risk of hypoglycemia.The computer-readable medium further includes instructions, when executed by the at least one processor, that cause the system to determine, in the insulin delivery monitoring unit, a future insulin target trajectory, determine, using the normative insulin planning unit, the amount of insulin needed to minimize the deviation determined by the model match assessment unit, adjust a proposed basal rate or a proposed bolus rate with the insulin need identified by the normative insulin planning unit to determine an approved basal rate or an approved bolus rate, and output the approved basal rate or the approved bolus rate.

It should be understood that the various techniques described herein may be implemented in connection with hardware or software components, or, where appropriate, with a combination of both. Exemplary types of hardware components that may be used include field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip systems (SOCs), complex programmable logic devices (CPLDs), and the like. The methods and apparatus of the presently disclosed subject matter, or certain aspects or portions thereof, may take the form of program code (i.e., instructions) embodied in a tangible medium, such as a floppy diskette, CD-ROM, hard drive, or any other machine-readable storage medium; when the program code is loaded into and executed by a machine, such as a computer, the machine becomes an apparatus for practicing the presently disclosed subject matter.

While example implementations may refer to utilizing aspects of the presently disclosed subject matter in the context of one or more stand-alone computer systems, the subject matter is not so limited and, rather, may be implemented in connection with any computing environment, such as a network or distributed computing environment. Furthermore, aspects of the subject matter disclosed herein may be implemented in or across multiple processing chips or devices, as well as stored across multiple devices. Such devices may include, for example, personal computers, network servers, and handheld devices.

Although the systems and techniques have been described in language specific to structural features and/or methodological acts, it should be understood that the systems and techniques defined in the appended claims are not necessarily limited to the specific features or acts described. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.

100 High-Level Functional Block Diagram 110 Insulin Device 120 Glucose Monitor 130 Processor 140 Patient 150 Activity Monitor 160 Smartphone 203 Setup Input Data 205 CGM Data 210 Other Sensed Input Data 215 Insulin Data 220 User Input Data 225 External Process Data 230 Insulin Delivery Rate Converter 235 Comparator 240 Glycemic Risk Assessment 245 Insulin Delivery Monitor 290 Output 420 State Estimator 425 Reference Insulin Rate (RIR)
430 Model Fit Evaluator 435 Deviation 440 Reference Insulin Rate (RIR) Updater 450 Internal Reference Insulin Rate 605 Input 620 Hyperglycemia Risk 640 Hypoglycemia Risk 820 Normative Insulin Planning 840 Monitoring 1000 Device 1002 Processing Unit 1004 Memory 1006 Dashed Line 1008 Removable Storage 1010 Non-Removable Storage 1012 Communication Connection 1014 Input Device 1016 Output Device

Claims (12)

1. A system comprising:
at least one processor;
A non-transitory computer-readable medium that, when executed by the at least one processor, provides the system with:
receiving input data including insulin data and glucose data at a comparison unit;
using a model agreement assessment unit of the comparison unit to identify a discrepancy between separately derived estimates of metabolic data derived from the insulin data and the glucose data by quantifying the extent to which recent blood glucose measurements are inconsistent with recent insulin, wherein the discrepancy is calculated as a difference between an estimate of the metabolic data made using an individualized physiological model and the input data and an estimate of the metabolic data made using the individualized physiological model and the input data without using the glucose data;
using a glycemic risk assessment component to quantify a level of future risk of hyperglycemia or hypoglycemia based on the predicted glucose data;
Using the insulin delivery monitor,
determining the amount of insulin required to minimize the deviation;
adjusting a suggested basal rate from an external source with the determined insulin need to determine an approved basal rate, the approved basal rate being calculated for a time window during which the required amount of insulin will be delivered, the time window being determined from the quantified level of risk;
outputting the approved basal rate;
a non-transitory computer readable medium comprising instructions for adjusting an insulin delivery rate based on data from the comparison component and the glycemic risk assessment component by
Including, the system. The system of claim 1 , wherein the insulin delivery monitor further considers a suggested bolus rate from an external process. 3. The system of claim 1 or 2, wherein the non-transitory computer-readable medium further comprises instructions that, when executed by the at least one processor, cause the system to use the insulin delivery monitor to convert the approved basal rate into a blend of a basal rate and individual boluses. 4. The system of claim 1, wherein the non-transitory computer-readable medium further comprises instructions that, when executed by the at least one processor, cause the system to determine a reference insulin rate (RIR) by a reference insulin rate update unit, the RIR being an internal reference for insulin that is considered to achieve equilibrium. 5. The system of claim 4, wherein the non-transitory computer-readable medium further comprises instructions that, when executed by the at least one processor, cause the system to receive the RIR at the glycemic risk assessor and use the RIR to quantify the risk of future hyperglycemia or future hypoglycemia. 6. The system of claim 4 or 5, wherein the non-transitory computer-readable medium further comprises instructions that, when executed by the at least one processor, cause the system to receive the RIR at the insulin delivery monitor and use the RIR to determine a future insulin target trajectory and an amount of insulin needed to minimize the deviation. 7. The system of claim 1, wherein the basal rate suggested by the external source is from a pre-programmed basal profile, patient-initiated insulin delivery, or an artificial pancreas algorithm. 8. The system of claim 1, wherein the non-transitory computer-readable medium further comprises instructions that, when executed by the at least one processor, cause the system to quantify a level of future hyperglycemic risk, wherein for a higher level of hyperglycemic risk, a shorter time window is determined as the time window during which the required amount of insulin will be delivered. The system of claim 1 , wherein the glucose data is received from a continuous glucose monitor. 10. The system of claim 1, wherein the insulin data includes previously administered basal/bolus insulin. The system of claim 1 , wherein the estimate of the metabolic data is estimated using a Kalman filter. The system of claim 1 , wherein the approved basal rate is output to an insulin delivery device.